Invar Resonator Housings: What Laser Manufacturers Specify
A resonator housing does one job: hold two or more mirrors, and often a gain medium, in a fixed geometric relationship so the light bouncing between them builds into a stable mode rather than walking off axis. The cavity length gets most of the attention in casual conversation about Invar and lasers, but the geometry that actually protects a laser’s output, mode quality and beam pointing over its service life is parallelism between mirror faces, concentricity of the bores that locate optics and the gain medium and flatness of the surfaces those parts mount against. This guide covers what a resonator drawing should call out and how the machining sequence keeps those callouts true after the part leaves the shop.
What a resonator drawing actually needs to specify
A cavity length dimension alone tells a shop almost nothing about what matters for laser performance. A complete resonator drawing typically specifies:
- Parallelism between the two (or more) mirror mounting faces, since a tilt here misaligns the beam path directly
- Concentricity of bores that locate mirrors, the gain medium or apertures on the optical axis, usually specified as total indicator reading
- Flatness of mating faces where an optic, a mount or a Peltier element sits, since a warped face tilts whatever is bolted to it
- Cavity length, the linear dimension between reference features, with its own tolerance
- Perpendicularity between the optical axis (the bore or mounting line) and each mirror face, closely related to parallelism but worth calling out separately when the housing has more than two optical elements
The optical bench worked example walks through a similar geometry problem for a bench holding several mounts rather than a resonator’s linear axis, and the general tolerance practice for Invar parts is covered in Invar machining tolerances.
Typical achievable callouts
The table gives typical, shop-dependent figures for a machined and, where noted, lapped Invar resonator housing. These are planning numbers. The achievable figure for your specific part depends on its span, its finish requirement and the shop’s process, and should be confirmed directly rather than assumed from this table.
| Feature | Typical machined tolerance | Typical with lapping or additional finishing |
|---|---|---|
| Mirror face parallelism | 15 to 30 arc seconds | 5 to 15 arc seconds |
| Bore concentricity (TIR) | 0.015 to 0.025 mm | 0.005 to 0.010 mm |
| Mounting face flatness | 0.005 to 0.015 mm | 0.001 to 0.005 mm |
| Cavity length | ±0.010 to ±0.025 mm | ±0.003 to ±0.008 mm |
Tighter numbers than these are achievable on specific features with enough process and inspection time. What is not achievable, at any tolerance, is holding these figures long term on a part that carries residual machining stress, which is the subject of the next section.
How the stress relief sequence protects the geometry
A resonator housing that measures perfectly on the day it ships and drifts out of parallelism six months into service has usually failed for the same reason: residual stress from machining released slowly after the part left the shop. The sequence that prevents this, described in general terms in stress relief for Invar machined parts, matters more on a resonator than on almost any other Invar part because the features at risk (parallelism, concentricity, flatness) are angular and relational, not just linear, and small angular shifts compound over the cavity’s length.
| Step | What it does for the geometry | Risk if skipped |
|---|---|---|
| Rough with even stock allowance | Leaves balanced material on both mirror faces and around each bore | Uneven stock locks in asymmetric stress that releases unevenly later |
| Stress relieve (typically 315 to 345 °C, 1 to 2 h) | Releases roughing stress before critical features are cut to size | Parallelism and concentricity drift weeks to months after shipment |
| Semi-finish, then finish with light cuts | Brings bores and faces to final size with minimal new cold work | New machining stress reintroduces the same drift risk after stress relief already addressed the old one |
| Stabilize, for the most demanding parts | Long low-temperature soak that settles the structure further | Slow, small drift continues over the part’s early service life |
| Inspect after a soak at 20 °C | Confirms the part’s actual room-temperature geometry, not a value distorted by handling warmth | A part measured warm or immediately off the machine can show numbers that do not hold once it settles |
Bores that carry the tightest concentricity requirement are usually drilled or bored undersize before stress relief and brought to final size only after it, for the same reason a precision bore in the optical bench example is finished last: any step performed before the part has released its stress risks being undone by that release later.
Choosing the process for the critical bore
Most resonator bores are finished by conventional boring, which gives good control over both diameter and concentricity and leaves a surface that is straightforward to inspect and, if needed, plate. Wire EDM is sometimes proposed for a bore in a thin wall or a hard-to-reach pocket, since it cuts without the clamping and cutting forces that can distort a delicate section, and Invar wire EDM machining covers where that trade-off makes sense. The catch is that EDM leaves a thin recast layer with different properties from the base metal, which is usually removed on a critical optical or sealing surface, adding a finishing step that a conventionally bored feature does not need. For most resonator housings with reasonable wall thickness, conventional boring after stress relief is the simpler and more predictable path, and EDM is reserved for geometries that boring genuinely cannot reach.
Worked example: a 150 mm cavity housing
This is a hypothetical illustration, not a real order or a real shop’s numbers. Take a resonator housing with a 150 mm cavity length between two mirror mounting faces, a parallelism requirement of 10 arc seconds between those faces and a bore concentricity requirement of 0.010 mm TIR for the gain medium.
Ten arc seconds over a 150 mm span corresponds to roughly 0.007 mm of height difference edge to edge on the mounting faces, which sits inside the lapped range in the table above but outside the plain machined range. That tells the shop the mirror faces need a lapping step after finish machining, not just a fine end mill pass. The 0.010 mm bore concentricity sits at the tighter end of the plain-machined range, achievable with a boring operation performed after stress relief rather than during roughing.
The shop’s proposed sequence: rough the housing with 0.5 mm of stock left on both mirror faces and the bore, stress relieve at 330 °C for 90 minutes, semi-finish the bore to 0.1 mm oversize, finish-bore to size, finish and lap the mirror faces to their parallelism callout, then inspect the whole part after a soak at 20 °C. This sequencing is what makes the parallelism and concentricity numbers on the drawing a real, lasting spec rather than a number the part happens to hit on day one.
Six months into service, a housing built this way should measure the same way it did at first article, within the normal soak-to-soak repeatability of the inspection method itself. A housing built without the stress relief step, even if it passed the same first article inspection, is the one more likely to show a slow parallelism shift as the roughing stress it never released continues to work its way out. The difference between the two housings is invisible at shipment and is exactly what the sequence in the table above is built to prevent.
The same reasoning applies if the housing later needs a design revision, for example a new port added for a monitor photodiode. Re-machining a feature into a housing that has already been stress relieved and stabilized should go through its own light stress relief afterward if the new cut removes a meaningful amount of material, rather than assuming the part’s earlier heat treatment still covers a geometry that has since changed. Treating a revision as “just adding one hole” on an otherwise finished Invar part is a common way a previously stable housing starts drifting again.
What to put on the resonator RFQ
Beyond the drawing’s geometric callouts, a resonator RFQ should state:
- Parallelism and concentricity tolerances explicitly, not just implied by a linear dimension, since these are the features that actually protect laser performance
- Whether lapping is expected on mirror mounting faces, and to what finish
- The stress relief cycle, either specified or required as a documented, fixed shop process
- Inspection temperature and soak time, and whether a CMM report is required per part, covered in Invar CMM inspection reports
- Bore finishing sequence, confirming precision bores are finished after stress relief, not before
Background on the alloy’s specification and typical properties is on the Invar specs reference page. The physics of why cavity geometry, not just length, governs a resonator’s output mode is covered in the Wikipedia entry on the optical cavity.
When you send an RFQ with your drawing, we match it with up to three shops that machine the nickel-iron alloys and can speak specifically to parallelism, concentricity and the stress relief sequence a resonator housing needs. This guide sits under the optics, photonics and laser systems overview with the other sourcing and process guides for that vertical.
Frequently asked questions
Why does parallelism matter more than the linear cavity length tolerance?
Cavity length sets the resonator's mode spacing, and a modest length error can often be tuned out elsewhere in the system. A parallelism error between the two mirror mounting faces tilts the beam path itself, which degrades mode quality and output power in a way that is much harder to correct after the fact. Both matter, but a drawing with a tight length tolerance and no parallelism callout is missing the harder requirement.
What arc second tolerance is realistic for a machined Invar mounting face?
Typical machined and lapped Invar mounting faces hold parallelism in the range of 5 to 15 arc seconds over a modest span, with tighter figures achievable with additional lapping and inspection time. Confirm the achievable number with the shop against your specific span and finish requirement rather than assuming a figure from another job transfers directly.
Does bore concentricity need to be held before or after stress relief?
Both, in stages. Bores are typically drilled or bored undersize before stress relief, brought closer to size afterward and finished to their final concentricity and size only once the part has released its roughing stress. Finishing a precision bore before stress relief risks the bore shifting when the part relieves stress later.
Can wire EDM hold the concentricity a resonator bore needs?
Wire EDM can hold tight concentricity and is useful for hard corners or thin sections, but it cuts without conventional machining force and leaves a recast layer that is usually removed on critical surfaces. For a bore that also needs a specific surface finish for an optical or sealing function, confirm with the shop whether EDM plus a finishing pass, or conventional boring, is the better fit for that specific feature.
Reviewed against
- ASTM F1684, Standard Specification for Iron-Nickel and Iron-Nickel-Cobalt Alloys for Low Thermal Expansion Applications
- Wikipedia, Optical cavity
Values in this guide are typical handbook figures for planning. The mill certificate for your heat lot and the shop's quote for your geometry govern.
Need a quote for this part?
Send the drawing. We match you with up to 3 shops that machine Invar and reply within 24 business hours.
Get matched with an Invar shopKeep reading
- A 300 by 200 mm optical bench in Invar, from alloy choice to quote · Choosing a Supplier
- Stress Relief for Invar Machined Parts: When and How Shops Do It · Process
- Reading an Invar CMM Inspection Report: What Each Number Means · Process
- Invar Machining Tolerances: What a Shop Can Hold and What It Costs · Process